How Do Judges Decide International Relocation Cases?

Rob George*

Abstract: There has been considerable controversy over the approach taken to international

relocation disputes by the courts of England and Wales, but little has been known about how

first instance judges approach such decisions in ‘everyday’ relocation cases. This article

reports the findings of an empirical study in 2012 which looks at case outcomes and judicial

reasoning in international relocation disputes based on the decisions of trial judges and the

reports of lawyers. The article analyses case characteristics and some factors which may

influence case outcomes, and explores the judicial reasoning used to reach those conclusions.

INTRODUCTION

For over a decade, international relocation disputes have remained one of English family

law’s more controversial issues. These difficult cases involve the application by one parent

for permission to remove their child permanently1 from the United Kingdom against the other

parent’s wishes. Until recently, the leading authority remained the 2001 Court of Appeal

decision in Payne v Payne,2 interpreted in the light of the 2011 Court of Appeal case of K v K

(Relocation: Shared Care Arrangement).3 The most recent decision, Re F (International

Relocation Cases),4 reduces the significance of Payne and calls for a holistic, case-specific

analysis.

While relocation law has been subjected to significant criticism,5 little is known about the

reality of these cases as they are being decided on the ground. The focus of scholars’

attention has understandably been on the principles as set out by appellate judges, together

with the decisions of those few reported first instance cases. Work which has asked about

everyday decision-making in relocation cases has approached the question indirectly, talking

* UCL Faculty of Laws.

This work was conducted when I was a British Academy Postdoctoral Fellow at University College, Oxford,

and I express my thanks to the Academy and to Oxford for their support. I also record my indebtedness to Sir

Mathew Thorpe, who was pivotal in making it possible to collect the data used in this article, along with his

then-Legal Secretaries Victoria Miller and Edward Bennett who did a lot of work to chase judgments for me.

Two sets of reviewers have provided detailed comments on this article, as have the Editors, and it has improved

considerably as a result.

1 Temporary relocation is an increasingly important area, but the issues are quite different from their

permanent relocation counterparts. 2 [2001] EWCA Civ 166, [2001] 1 FLR 1052 (hereafter, ‘Payne’).

3 [2011] EWCA Civ 793, [2012] 2 FLR 880 (hereafter, ‘K v K’).

4 [2015] EWCA Civ 882 (hereafter, ‘Re F (International Relocation Cases)’).

5 See, eg, M. Hayes, ‘Relocation Cases: Is The Court of Appeal Applying the Correct Principles?’ [2006]

CFLQ 351; J. Herring and R. Taylor, ‘Relocating Relocation’ [2006] CFLQ 517; F. Judd QC and R.

George, ‘International Relocation: Do We Stand Alone?’ [2010] Family Law 63; S. Gilmore, ‘The Payne

Saga: Precedent and Family Law Cases’ [2011] Family Law 970; R. George, ‘Reviewing Relocation?’

[2012] CFLQ 110; R. George, Relocation Disputes: Law and Practice in England and New Zealand

(Oxford, Hart Publishing, 2014) (hereafter, ‘Relocation Disputes’).

to practitioners about their experiences.6 This work is valuable, not least in understanding the

wider processes involved in family dispute resolution of which, of course, the family court is

only part. However, such research inevitably gives a somewhat detached perspective, asking

judges and practitioners in the abstract rather than examining actual decisions.

At the same time, a major debate in England and Wales about relocation law is whether leave

to remove a child from the jurisdiction is granted too easily. This important debate has been

significantly hampered by a lack of knowledge about the proportion of applications that is, in

fact, allowed. More generally, little is known about the people involved in relocation

disputes, such as where they are seeking to go and why, how old their children are, and so on.

This paper reports the findings of empirical work carried out in 2012 designed to shed light

on these issues. In particular, the following questions are addressed:

(1) What are the characteristics of international relocation cases decided by courts in

England and Wales, in terms of the families involved and case details?

(2) What are the outcomes of litigated cases in terms of applications being allowed or

refused, and what variables appear to influence those outcomes?

(3) How are judgments constructed in international relocation cases to allow judges to

reach these conclusions?

INTERNATIONAL RELOCATION: THE LEGAL FRAMEWORK

As with all private law disputes about children’s upbringing, the welfare of the child is the

court’s paramount consideration in a relocation dispute,7 supplemented by the welfare

checklist in s 1(3) of the Children Act 1989. In addition, when the cases being discussed in

this article were decided, the courts remained focused on the guidance from Payne, which

highlighted factors of particular relevance to the relocation context. In Thorpe LJ’s judgment,

a so-called ‘discipline’ was set out:8 the discipline asked about the applicant’s motivation and

the level of planning that had gone into the application, followed by the respondent’s

motivations, and then the effect of a refusal of the application on the applicant, all of which

was said to be relevant to an overall analysis of welfare. Similar guidance was given by the

then-President of the Family Division, Dame Elizabeth Butler-Sloss,9 though in the years

following Payne it was Thorpe LJ’s judgment which dominated judicial thinking.

The Payne guidance, particularly Thorpe LJ’s ‘discipline’, was once ‘commonly seen as

walking and talking like a presumption’ in favour of relocation applications being allowed.10

6 Relocation Disputes.

7 Children Act 1989, s 1.

8 Payne, [40]–[41].

9 Payne, [85]-[86].

10 C. Geekie QC, ‘Relocation and Shared Residence: One Route or Two?’ [2008] Family Law 446, 151–52;

see also M. Hayes, ‘Relocation Cases: Is the Court of Appeal Applying the Correct Principles?’ [2006]

CFLQ 351; Relocation Disputes, pp 88-90 and 90-112.

However, more recent decisions – especially K v K (Relocation: Shared Care

Arrangement),11

which was decided shortly before the cases reported in this research were

decided, and Re F (Relocation),12

which was decided three quarters of the way through the

sampling period for this research – made clear that Payne was intended to guide judges in the

factors which they might need to consider, but not as to the outcomes they should reach.13

In

practical terms, this was generally thought to mean that the ‘rigour’ with which judges are

approaching applications increased: the impression that a relocation application is almost

bound to succeed was dying away in 2012 when this research was conducted.14

Consequently, although the guidance itself did not change, it was thought that the way in

which judges approached cases using the guidance did.

The approach to international relocation cases has subsequently shifted again, with the 2015

Court of Appeal decision in Re F (International Relocation Cases).15

Here, the appeal court

said for the first time that following Payne’s guidance may amount to an error of law if the

facts of the case being decided call for a broader (or simply different) approach; the count

commended a global, holistic analysis of all factors, and a parallel analysis of the various

proposals being put forward in an assessment of welfare. While this approach draws heavily

on K v K, it marks a change of emphasis from the position as understood before mid-2015;16

when thinking about the cases discussed in this article, it is therefore important to keep in

mind the legal landscape at the time, and how it appears to have shifted subsequently.

SUMMARY OF METHODOOGY

This paper reports findings from two related studies.17

The first study, which is the principal

source of data, comprises 96 judicial decisions in international relocation cases in the

calendar year 2012, and referred to as the Court Case (‘CC’) sample. The CC sample

comprises 96 international relocation decisions where either the court’s order (N=43), the

court’s judgment (N=53), or both (N=20) were provided. For obvious reasons, judgments

provide a great deal more information than orders alone, though most orders submitted

contained detailed recitals and other information, and even the most basic usually contained

considerable information. Nonetheless, it is worth noting that in the cases where only the

order was available, less information was available and so the analysis that can be performed

11

[2011] EWCA Civ 793, [2012] 2 FLR 880. 12

[2012] EWCA Civ 1364, [2013] 1 FLR 645. 13

See, eg, Re H (Leave to Remove) [2010] EWCA Civ 915, [2010] 2 FLR 1875; J v S (Leave to Remove)

[2010] EWHC 2098 (Fam), [2011] 1 FLR 1694; Re W (Relocation: Removal Outside Jurisdiction)

[2011] EWCA Civ 345, [2011] 2 FLR 409. 14

Relocation Disputes, p 109. 15

[2015] EWCA Civ 882. 16

E Devereux and R George, ‘“Alas Poor Payne, I knew him…” An Interpretation of the Court of Appeal’s

Decision in Re F (International Relocation Cases) [2015] EWCA Civ 882’ [2015] Family Law, October .

For a summary of the law at the time of this research, see R George, F Judd QC, D Garrido and A

Worwood, Relocation: A Practical Guide, 1st Edn (Jordans, 2013), Ch 2. 17

A full methodological appendix, highlighting various limitations of the dataset, is included at the end of

this paper.

on those decisions is more limited. All 96 cases contribute to the quantitative part of this

study, and the 53 cases where judgments were provided underpin the qualitative part.

The second study, designed to augment and test the reliability of the CC sample, comes from

52 questionnaire responses from family lawyers in England and Wales between October 2012

and January 2013, and referred to as the Research Questionnaire (‘RQ’) sample. These 52

cases augment the quantitative part of this paper, but not the qualitative.

Because the CC and RQ samples ask about cases decided in the same period, there is obvious

scope for overlap. Cases were analysed for overlaps, looking at key characteristics such as

number and age of children and proposed destination country. Nine likely overlap cases were

identified in this way.18

When the two samples are analysed separately, all cases are kept in

each sample; however, in places the two samples are combined and then the nine overlap

cases are included only once. Where this is done, the CC cases are used, since the CC sample

is more reliable: the data come directly from court documents, whereas the RQ sample relies

on practitioners’ recollection, potentially some months after their involvement with the case

ended.

QUANTITATIVE FINDINGS

Case Outcomes

England has traditionally been seen as a ‘pro-relocation’ jurisdiction, with applications to

remove a child from the jurisdiction perceived to be treated comparatively favourably.19

However, while some other countries have researched case outcomes to test this question, no

work has previously been done in England and Wales, so making it difficult to know whether

the perception is reflected in practice. The position in Canada (perceived as ‘neutral’ when it

comes to relocation applications) and New Zealand (perceived as ‘anti-relocation’) make for

useful broad-brush comparators, though the different methodologies used for collecting cases

in different studies mean that caution is needed in making direct comparison. In Canada,

Rollie Thompson’s research (which involved only reported cases, though a high proportion

are reported in Canada) showed an overall success rate of 68% for international applications

going outside North America.20

Similarly, Mark Henaghan’s New Zealand study (which

involved all first instance decisions in that jurisdiction) also showed an overall success rate of

68% for international cases in the courts in that country.21

Given that these jurisdictions are

18

A summary of all cases in the two samples, which indicates the likely overlap cases, is contained in R.

George, Relocation Disputes in England and Wales: First Findings from the 2012 Study (Oxford Legal

Research Paper 91/2013), online at ssrn.com/abstract=2306097, Appendix 2 (hereafter, ‘First Findings’) 19

T. Foley, International Child Relocation: Varying Approaches Among Member States to the 1980 Hague

Convention (London, Court of Appeal, 2006); Relocation Disputes, pp 10-12. 20

R. Thompson ‘Heading for the Light: International Relocation from Canada’ (2011) 30 Canadian Family

Law Quarterly 1. Many moves to North America from Canada involve very short distances just across

the border, which may be why differentiating broadly between North American and other destinations is

helpful. 21

M. Henaghan, ‘The Conundrums of Relocation’ (London Metropolitan University Conference, July 2013).

said to represent different approaches to relocation, it is noteworthy that the reported success

rates of applications appear so similar.

The two datasets in this study yield slightly different findings from each other in terms of

overall outcomes: the CC sample had an overall success rate of 71.9%, the RQ cases 59.6%;

when the two datasets are combined (and the supposed duplicates discounted), the overall

success rate is 66.7%. Whichever percentage is used, the comparison with the results of the

(broadly comparable) studies in other jurisdictions, as shown in Table 1, is interesting:

apparently challenging the standard depictions of these jurisdictions as respectively pro-,

neutral, and anti- regarding relocation, the findings show little difference in these broad terms

between the court outcomes in the three jurisdictions.22

Table 1: Table showing the percentage of international relocation applications that

were allowed in the CC and RQ datasets, and in studies from Canada and New

Zealand

Data Set N Percentage of Cases where

Relocation Is Allowed

INT CC

96 71.9

INT RQ, Judicially Determined Cases Only

52 59.6

INT CC and RQ Combined

139 66.7

International Relocation Cases in Canada,

2005-10 (‘neutral’) 47 68.0

International Relocation Cases in New

Zealand, 2011-12 (‘anti-relocation’) 44 68.0

Of course, the results of court litigation do not tell the whole story, because cases which reach

the courts are filtered in various ways by pre-litigation advice and decision-making, all of

which (it is generally assumed) takes place in the shadow of the law,23

influenced by people’s

perceptions of the law as well as non-legal factors and individual bargaining chips. In the

relocation context, earlier research has suggested that the advice that lawyers give to potential

applicant and respondent parents varies considerably from country to country,24

and this is

likely to affect the overall picture of relocation case outcomes. On the other hand, the

proportion of international relocation cases which settle is thought to be relatively low.25

22

None of the differences is statistically significant, whichever of the CC, RQ or combined results is used

for comparison with the Canadian and New Zealand data.

23 R. Mnookin and L. Kornhauser, ‘Bargaining in the Shadow of the Law: The Case of Divorce’ (1979) 88

Yale Law Journal 950 24

Relocation Disputes, esp pp. 85-86. 25

Ibid, p. 3, reporting the views of lawyers in England and New Zealand, both of whom said that it was

unusual to settle relocation cases.

While it is interesting and important to know the overall success rates for international

relocation cases in this sample, the next question is to look at the factors which affect those

case outcomes. It is not the case that every relocation application has an equal chance of

succeeding. Practitioners have long argued that the motivation of the applicant probably is,

and certainly should be, a factor influencing whether the relocation should be allowed or

not;26

and more recently, the English courts have placed renewed focus on the importance of

the child’s existing care arrangements before the relocation dispute begins as a factor which

may influence the decision about relocation.27

Consequently, the analysis now moves to look

at factors which, from a statistical perspective, appear to be relevant to the decision-making

which leads to the overall outcomes reported above.

Factors Influencing Case Outcomes

The study approaches this analysis using different methodological techniques. The first,

called an Information Gain Score (IGS) analysis, ranks each of the variables according to

how much additional information that variable provides when predicting outcomes. The

results of that analysis are reported in an earlier paper,28

their main function being to guide

further work using other methods. The findings of that further work are set out here, divided

into two broad sections.

Families and biographical characteristics

First, we can note that, in line with anecdotal reports, the vast majority of applicants in this

study were mothers. In the CC sample, 93% of applicants were mothers, 5% were fathers,

and 2% were grandparents; in the RQ sample, 95% were mothers and 5% fathers. This

pattern is broadly similar to that in other countries; for example, a Canadian study of

relocation decisions from 2001 to 2010 showed 92% of applicants were mothers, 7% were

fathers and 1% were non-parents.29

One of the most complete pieces of data available in the study is the children’s ages. In the

CC sample (N=140 children), the mean age was 6.75 years, while in the RQ sample (N=88

children) it was 7.9 years. The spread of ages is illustrated in Figure 1, where the percentage

of children of each age in each sample is shown on a bar chart (with percentages used for

ease of comparison). It can be seen that while the RQ sample has a single peak at 8 years, the

CC sample has a higher percentage of children of pre-school age. It is not known why the two

samples would be different in this respect.30

26

See, for example, the argument of Eleanor Hamilton QC, as she then was, in Re B (Leave to Remove:

Impact of Refusal) [2004] EWCA Civ 956, [2005] 2 FLR 239; see also Relocation Disputes, Ch 4. 27

See especially K v K. 28

‘First Findings’, paras 4.9-4.18. 29

N Bala and A Wheeler, ‘Canadian Relocation Cases: Heading Towards Guidelines’ [2012] Canadian

Family Law Quarterly 271, 289. 30

In ‘First Findings’, paras 3.27-3.35, this issue is explored in more detail.

It is also notable that in both samples, children over the age of about 10 or 11 feature

relatively infrequently. When we look later at the qualitative analysis of court judgments, one

reason why this might be the case is that older children’s views are more influential in

decision-making, and so parents may settle litigation (or not start it at all) in some cases

involving an older child, once the child’s views are known.

However, that possible pre-litigation filter issue aside, the data do not support the suggestion

that the children’s ages might affect case outcomes. Indeed, case outcomes are stable

whatever the age of the child involved.31

Figure 1: Bar chart showing the percentage of children of each age in each of the CC

and RQ samples

Another issue to consider is the care arrangements in place for children prior to the relocation

application. One might expect to find that the more the child’s day-to-day care was shared by

the respondent parent, the less likely it would be that a relocation application would succeed.

The intellectual justification for this approach would be that such a child has ‘more to lose’

by relocating, in terms of his or her relationship with the respondent parent, but also that an

existing shared care arrangement would allow the child more readily to remain in the present

location in the main care of the respondent parent if the applicant parent decided to relocate

31

‘First Findings’, paras 4.51-4.55.

0

2

4

6

8

10

12

14

16

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

Pe

rcen

tage

of

Child

ren

Age of Child

CC Cases (N=140) RQ Cases (N=88)

anyway.32

Table 2 shows the success rates for international relocation applications in the two data

samples.33

Cases were classified as ‘shared care’ where the child spent at least 35% of time

with each parent, ‘overnight stays’ where there was overnight time but less than 35%, ‘direct

contact’ where there was no (or no regular) overnight time, and ‘no direct contact’ where the

child had no face-to-face time with a parent; there was also an ‘other’ category, for cases

where (for example) a non-parent was also involved in the child’s main care, or where the

respondent parent was the main carer.34

In the CC sample, this information was assessed by

the researchers either directly or indirectly from the court documents; in the RQ sample,

respondents reported the care arrangements that had been in place. It can be seen that in the

CC sample, the success rates of applications follow the approximate pattern that would be

expected, with shared care cases least likely be allowed and no contact cases most likely to

be. However, the RQ sample shows almost no difference between the main categories of care

arrangement, which may lead to more doubt about the overall picture. In any case, the

differences are not statistically significant in either dataset, so the differences seen may

simply be the result of random chance.

32

An alternative way of thinking about this issue – in favour of relocation – is to say that a child in a shared

care arrangement has a stronger existing bond with the respondent parent, and so will be better placed to

maintain a positive and beneficial relationship after relocation. 33

Information about care arrangements is not available for all cases in either sample. 34

This classification was not used consistently by judges in their judgments, with the term ‘shared care’ used in

many cases involving overnight time with both parents regardless of the extent of sharing.

Table 2: Table showing success rates for international relocation applications in the

CC and RQ samples, and in the two combined, for different pre-relocation child care

arrangements

Care

arrangement

CC Cases RQ Cases CC and RQ Combined

N Success

rate (%) N

Success

rate (%) N

Success

rate (%)

All cases

where care

arrangements

known

71 66.2 51 60.8 118 61.9

Shared care

arrangement

(35% <)

8 50.0 16 56.3 24 54.2

Some

overnight

stays

43 60.5 22 50.0 63 55.6

Direct contact,

no overnight

stays 11 72.7 6 50.0 16 62.5

No direct

contact 5 100.0 3 100.0 8 100.0

Other care

arrangements 4 100.0 4 100.0 7 100.0

That caveat – given the inevitable limitations of the dataset – aside, these findings might

incline us to suggest that cases involving no overnight contact with the respondent parent are

more likely to be allowed than those involving at least some overnight stays. However, it is

more difficult to know whether, once that line has been crossed, the amount of sharing of

care makes any difference: the CC data suggest that it does, but the RQ and combined data

are more equivocal. Further work is needed to clarify this issue. Interestingly, as noted below

in the court judgments analysis,35

judges seldom varied their language in describing the

relationship between the child and the non-moving parent. Save where that parent presented

some kind of risk to the child, they were commonly described as, for example, having ‘a very

loving and close relationship’, whatever type of care arrangement was in place.

A separate issue which the data seem to suggest may be of some relevance is the applicant

parent’s current relationship status, whether married, cohabiting, and so on.36

As the figures

in Table 3 show, it seems that in this sample those applicants who were in some kind of

stable relationship had a higher chance of their applications succeeding than those who were

35

Below, text after fn 81. 36

There is no apparent variation in relation to respondent parents’ relationship status, as might be expected:

‘First Findings’, para 4.57.

not in an on-going relationship.37

Table 3: Table showing success rates for international relocation applications in the CC

sample for different current relationship statuses of the applicant parent

Current relationship type

APPLICANT PARENTS

N Success rate (%)

All cases where relationship status

known 47 68.1

Married

17 82.4

Engaged

2 [100.0]

Cohabiting

4 [75.0]

Non-cohabiting relationship 4 [75.0]

No on-going relationship 20 50.0

[ ] = small numbers, therefore percentages should be interpreted with caution

Looking in more detail at the applicants who are married or engaged compared with those

who are not in an on-going relationship, some striking differences emerge. For example,

amongst the 19 single applicants from whom this information is known, none was a British

national seeking to emigrate, whereas 14 of the 19 married or engaged applicants were.

Similarly, the vast majority of single applicants were seeking to relocate in order to return

home (80%) or to get increased family support (80%); between them, these reasons

accounted for nine out of ten cases involving single applicants. Of the remaining two cases,

one was based on a job offer abroad,38

while the other was seen as a pure ‘lifestyle’

decision.39

Amongst the married applicants, on the other hand, the clear majority were bringing an

application based primarily on a factor related to their new spouse. In some cases, he was

already in the new location (N=6), while in other cases he was seeking to return to his

original home country (N=5). Four of these eleven cases also involved the new partner

37

The result for those applicants not in an on-going relationship is statistically significant at the 5%

confidence interval, with a p-value of 0.022. 38

Case CC-81. The mother was offered a new job with her existing employer in its Brussels office, having

moved to London from another EU country in 2007. 39

Case CC-75. Note, though, that a background of domestic violence and child abuse by the father made this

case atypical as ‘lifestyle’ applications go.

having a job offer, and job offers for the partner were the main basis for a further three cases.

A few applications amongst the married group were based on the applicant herself seeking to

return to her original home country (N=3) or were pure ‘lifestyle’ cases (N=2).

It is interesting to wonder how best to understand these patterns. Are judges concerned to

protect applicants’ new relationships, which some authorities suggest will struggle if a new

partner is prevented from pursuing his life ambitions involving relocation, so as to try to

avoid the children experiencing a second relationship breakdown?40

Do some judges take the

view that children need a father-figure in their lives, and so keep children of single applicants

near the respondent fathers while allowing re-partnered applicants to move because this role

is seen as being performed by the step-father?41

Alternatively, is it that judges are generally

less impressed by applications premised on a desire by the applicant to return ‘home’, but

find those cases motivated by new partners or their careers more compelling?

Curiously, this aspect does not feature strongly in judicial reasoning. Judges frequently talked

about the effect of a refusal of the application, but rarely linked this discussion to issues

relating to new partners. As we will see,42

the focus was more often on the applicant’s present

life in this jurisdiction and, in many cases, a need for support in the proposed destination

country, neither of which tended to be linked to new partners.

Relocation Destinations and Motivations

The second broad category of findings relates to the destinations of the proposed relocations,

and the parties’ motivations in seeking or opposing such a move. Starting with proposed

relocation destinations, most international applications involve to one of three groups of

countries: countries within the EU, countries in North America, and countries in Australasia;

for the purposes of this analysis, other destinations are treated together as a collective ‘other’

group. Table 4 shows the success rates for applications in these four groups for the CC and

RQ data sets, and for the two combined.

40

See, for example, Re B (Removal from Jurisdiction); Re S (Removal from Jurisdiction) [2003] EWCA Civ

1149, [2003] 2 FLR 1043, [11]-[12] and [31]-[32]. 41

See generally R. van Krieken, ‘The “Best Interests of the Child” on Parental Separation: On the

“Civilizing of Parents”’ (2005) 68 Modern Law Review 25, 30-31 and 35. 42

Below, text after fn Error! Bookmark not defined..

Table 4: Table showing success rates for relocation applications in the CC and RQ

samples, and in the two combined, for different proposed destinations

Proposed

relocation

destination

CC Cases RQ Cases CC and RQ Combined

N Success

rate (%) N

Success

rate (%) N

Success

rate (%)

All cases

where

destination

known

95 72.6 51 60.8 139 67.6

European

Union 35 80.0 20 50.0 54 68.5

North

America 21 71.4 13 61.5 31 64.5

Australia /

New

Zealand

23 52.2 10 [60.0] 32 56.3

Other

destinations 16 87.5 8 [87.5] 22 86.4

[ ] = small values, therefore percentages should be interpreted with caution.

The CC data, on the left side, appear to tell a fairly coherent story: put simply, the further

away the proposed destination, the less likely it was that the relocation would be allowed.

The RQ cases, on the other hand, tell a different and in many ways less clear story. Proposed

moves to the EU here had the lowest success rate (50.0%), and there is little difference

between North American cases (61.5%) and Australasian cases (60.0%). The big variation in

outcomes for EU cases between the CC and RQ data is particularly puzzling, since this group

is the largest in both datasets and ought, therefore, to be least vulnerable to random

fluctuations. Looking deeper, though, there may be reason to think that the RQ sample of EU

moves is populated with ‘unusual’ cases. For example, while just 5.7% of the CC sample of

EU moves (N=35) were heard in the High Court (either by full time Judges or by Deputies),

almost a third (31.3%) of RQ cases to the EU were heard in that court (N=16). Consequently,

there may be reasons to think that the CC sample is more ‘typical’,43

and so to give greater

weight to the higher success rate for EU applications seen in that data. This interpretation also

fits with a more natural explanation of the patterns in the data, since it makes more sense if

comparatively short-distance moves within the EU should be allowed more frequently than

proposed relocations to the other side of the world.44

Turning to applicants’ reasons for seeking to relocate, earlier research in other countries has

43

Few cases are heard by High Court Judges, and almost all of them involve moves outside the EU unless

they have unusual complexities. For a third of all EU cases to be heard by High Court Judges suggests a

sample of cases which were unusual or complex on their facts, possibly caused by research participants

more readily remembering an ‘interesting’ High Court case than a ‘normal’ case before a lower level of

judge. 44

Conversely, language and cultural differences may be less with a long-distance Commonwealth country than

some short-distance European ones where the applicant and children are originally British nationals – but that

may be less of a factor where applicants want to return to their original home country.

shown that most applicants give several reasons.45

Consequently, while we can look, for

example, at all cases where the applicant was seeking to ‘return home’ to her original

country, and at cases which involved a move for a new job, some cases will fall into both

groups.

Table 5 breaks down cases by reference to the three main reasons given: returning home,

moving to take up a new job that the applicant or her new partner had been offered, or

moving for a generally improved lifestyle; plus a catch-all ‘other reasons’ category. Since the

categories are not mutually exclusive, the total number of cases for each dataset is less than

the sum of each reason for moving.

Table 5: Table showing success rates for international relocation applications in the

CC and RQ samples, and in the two combined, for different primarily reasons given by

the applicant for seeking to relocate

Reason for

seeking

relocation

CC Cases RQ Cases CC and RQ Combined

N Success

rate (%) N

Success

rate (%) N

Success

rate (%)

All cases

with known

reasons

75 72.0 52 59.6 123 65.9

Return to

applicant’s

original

country

42 73.8 23 73.9 61 72.1

New job for

applicant or

partner

16 93.8 25 56.0 40 70.0

Lifestyle

choice 21 47.6 21 52.4 40 47.5

Other

reasons 9 [66.7] 5 [40.0] 14 57.1

[ ] = small values, therefore percentages should be interpreted with caution

NB: Categories of reasons for seeking relocation are not mutually exclusive, therefore the sum of all reasons is greater than

the total number of cases.

These data again suggest a somewhat confused picture between the two datasets, with the CC

data suggesting that moving for a new job has a significantly higher chance of success while

moving to go home is about average, whereas the RQ data show the exact opposite. Both

datasets show that lifestyle choice cases are less likely to succeed than any other reason,

though, and this finding is statistically significant in both the court case and combined

datasets. This finding is consistent with practitioners’ views from earlier research suggesting

45

See N. Taylor, M. Gollop and M. Henaghan, Relocation Following Parental Separation: The Welfare and

Best Interests of Children (Dunedin, Centre for Research on Children and Families, 2010), p. 26; R.

Kaspiew, J. Behrens and B. Smyth, ‘Relocation Disputes in Separated Families Prior to the 2006

Reforms: An Empirical Study’ (2011) 86 Family Matters 72, 74; P. Parkinson, J. Cashmore and J.

Single, ‘Mothers Wishing to Relocate with Children: Actual and Perceived Reasons’ (2011) 27 Canadian

Family Law Journal 11.

that ‘lifestyle choice cases are the hardest to [succeed in bringing]’.46

Finally, we can look briefly at the way in which some of these factors appear to interact.47

Most interesting is the interaction between the proposed destination and the applicant’s

principal reason for seeking to relocate, as set out in Table 6.

Table 6: Table showing success rates for relocation applications in the CC sample,

divided into the main geographic locations of the proposed destinations, for different

primarily reasons given by the applicant for seeking to relocate

Reason for

seeking

relocation

European Union

North America Australia / New

Zealand

N= Success

rate (%) N=

Success

rate (%) N=

Success

rate (%)

All cases

where reasons

known

28 78.6 16 68.8 18 50.0

Return to

applicant’s

original home

country

18 83.3 10 70.0 7 [42.9]

New job for

applicant or

partner

5 [100.0] 2 [100.0] 5 [80.0]

Lifestyle

choice 6 [33.3] 5 [80.0] 10 40.0

[ ] = small values, therefore percentages should be interpreted with caution. NB: Categories of reasons for

seeking relocation are not mutually exclusive, therefore the sum of all reasons will be greater than the total

number of cases.

The pattern is not entirely consistent between the three groups of destinations, and the small

numbers in many of the categories mean that significant caution is needed when looking at

many of the percentages. But the patterns amongst these cases are largely unsurprising, given

the previous data seen. For example, permission to move to Australia in ‘lifestyle’ cases was

particularly unusual, whereas ‘going home’ to EU countries was very often allowed.

QUALITATIVE ANALYSIS OF COURT JUDGMENTS

So far, the analysis in this article has focused primarily on individual variables and how they

may impact on case outcomes. In practice, of course, each case involves many variables

which interact. As noted at the end of the previous section, some multi-variable analysis can

be done in a statistical way, but that analysis can inevitably only go so far, because the

46

Relocation Disputes, p. 102. 47

It is not possible to do this for all variables, as the data are not clear enough to lead to robust conclusions.

outcomes of court cases depend upon considerations which simply cannot be quantified. If

we knew that factors x, y and z added up led to outcome 1, and factors a, b and c added up led

to outcome 2, there would be little need for courts at all. The reality is that many of the

considerations which inform judicial reasoning and hence court outcomes are not quantifiable

in this way. In particular, the personalities of the people involved often have an effect – even

a decisive effect – on the judge’s decision.48

Consequently, we turn now to look at judges’

reasoning as expressed in judgments as a way better to understand how various factors fit

together in leading to case outcomes.

In looking at the 53 judgments available for this analysis, a number of themes can be

identified. First, it is interesting to see how the judges constructed the law itself, particularly

as this issue is particularly difficult and, perhaps, controversial following K v K.49

From there,

we turn to look at particular issues raised by the Payne guidance which are of especial

relevance to the overall reasoning processes seen in the judgments.

Constructions of the law

The treatment of the Payne guidance

In general, judges did not engage in lengthy discussion of the law relevant to relocation cases.

Most cases quoted the key summaries from Payne,50

and many judges proceeded to address

each of the points raised using headings based on either Thorpe LJ’s ‘discipline’ or Butler-

Sloss P’s ‘guidance’ (or sometimes both). Other authorities were mentioned intermittently.51

Most judges also gave express consideration to the items on the welfare checklist.52

While judges frequently reiterated that ‘[i]t is imperative ... that the court heeds the words of

caution of the Court of Appeal in K v K and avoids the risk of elevating the guidance in

Payne to the status of principle or, worse still, presumption’,53

only occasionally did they

make evaluative comments about the authorities. One District Judge discussed at some length

various criticisms of Payne before somewhat begrudgingly saying that ‘I shall ... submit to

the “discipline” suggested by Lord Justice Thorpe’.54

The only point on which there was any real variation in approaches was about cases

48

Relocation Disputes, p 173. 49

See, eg, E. Devereux and R. George, ‘When Will the Supreme Court Put Us Out of Our Payne?’ [2014]

Family Law 1586. 50

Payne v Payne, [40]-[41] and [85]-[86]. 51

The most commonly cited was K v K, followed by J v S (Leave to Remove) [2010] EWHC 2098 (Fam),

[2011] 1 FLR 1694 and Re AR (A Child: Relocation) [2010] EWHC 1346 (Fam), [2010] 2 FLR 1577; in

the final months of 2012, judges also sometimes cited Re F (Relocation) [2012] EWCA Civ 1364, [2013]

1 FLR 645, which was decided in October 2012. 52

Children Act 1989, s 1(3). 53

CC-34, [86]. 54

CC-46, [27]; also CC-58, [7].

involving shared care arrangements, which was described by one judge as a ‘particularly

vexed question’.55

There were two clear views in the eight cases where this issue was

discussed.56

One was that the court’s approach would be different in a shared care

application, and therefore that ‘[w]hether or not this is a shared care case is a question, in my

judgment, which I need to deal with at the outset of any discussion of the relevant law’.57

For

these judges, a finding of shared care meant that the Payne guidance was often given little

further attention, whereas if the judge declined to label the arrangements as such then Payne

was the court’s focus. It is interesting to wonder whether some judges might have used this

labelling as a means of reinforcing a decision already made on another basis: for example, if

a judge were inclined to refuse a relocation application for other reasons then categorising the

case as ‘shared care’ might help to bolster that conclusion if Payne is seen as not applying to

shared care cases. There is no way to assess whether this is so from court judgments, but

earlier work suggests that judges may be conscious of these considerations.58

Other judges rejected the idea that shared care cases called for a different legal approach –

‘the court should not categorise cases in accordance with the concepts of primary or shared

care’.59

In a clear reflection of the arguments put forward by counsel, judges in this group

often commented that they were ‘not assisted ... by attempting to decide who was the primary

carer’.60

These judges avoided any classification, but were happy to state (as Black LJ had

done in K v K) that ‘the guidance set out in Payne carries less weight’ in cases involving truly

shared care.61

It is likely that this confusion over shared care relocation cases has been resolved by the

October 2012 decision in Re F (Relocation),62

wherein Munby LJ reiterated the non-

classificatory approach and declared that Moore-Bick and Black LJJ represented the majority

in K v K,63

and more so by Ryder LJ’s decision in Re F (International Relocation Cases).

However, it is clear that throughout 2012 there was variation in approaches to these cases,

and judges felt compelled to engage with these questions.

Human rights law

55

CC-81, [104]. 56

Not all of the cases discussing this issue were classified as shared care cases – judges sometimes discussed

the controversy but then determined that the facts of the case fell outside the shared care area.

Consequently, the 8 cases here do not correspond exactly with the 8 shared care cases shown in Table 2;

in 3 of those cases, only an order was sent to the study, so no comment can be made about the judges’

approaches in their judgments. 57

CC-23, [42]; also CC-27, [7]; CC-12, [33] and CC-54, [29]. 58

Relocation Disputes, pp 69-71 and 78-79. 59

CC-96, [5]. 60

CC-61, [36]. 61

CC-67, [19]. 62

[2012] EWCA Civ 1364, [2013] 1 FLR 645 63

On the confusion over the ratio of K v K, see R. George, ‘Reviewing Relocation?’ [2012] CFLQ 110.

Another legal point, conspicuous primarily by its absence, was discussion of human rights

issues. Very few cases even mentioned the rights of the children, parents or others involved;

those that did contained only perfunctory comments. The handful of cases where rights were

mentioned fell into two groups. In one group, the mention of rights involved a purely abstract

acknowledgement that the parties had rights which were engaged by the decision:

I am also acutely conscious of each of the parties’ Article 6 and Article 8 rights under the

human Rights Act and, in coming to my decision, I have all of that at the forefront of my

mind.64

The other group of cases involved judges going one step further and noting (still only in the

abstract) that any interference with Article 8 rights needs to be necessary and proportionate

and in the pursuit of a legitimate aim. Two cases – which mark the high point of the rights

analysis in the relocation judgments – put the issue in these ways:

This order does affect each party’s rights to family life under the European Convention.

The interference in the father’s is merited by the welfare interests of his son and is

proportionate.65

[In allowing relocation] I would interfere with the father’s right to family life pursuant to

Article 8 ECHR, which I can only do if I consider it necessary and proportionate. I do

consider it necessary and proportionate to do so.66

It would be easy to criticise these judgments for their minimal engagement with the

potentially substantial interferences with the rights of the parties involved. But these judges at

least commented expressly, albeit minimally, on the human rights issues in their judgments.

In practice in the relocation cases in this study, therefore, human rights issues were not a

central focus for judges. However, human rights issues have a potentially substantial effect

on analysis of the issues.67

It is not that judges should undertake a rights analysis as a sort of

pro forma, but their analysis of ‘the entire family situation’ needs to demonstrate that they

have engaged in an overarching balancing of interests.68

A mere acknowledgement of rights

or assertion of proportionality may not advance matters much, and may not be adequate to

allow an appeal court to review the judge’s ‘compliance or otherwise’ with human rights

obligations.69

That said, although the Court of Appeal in Re Y suggested that no separate

analysis of Article 8 is required in private law children cases,70

in Re F (International

Relocation Cases) Ryder LJ held that in a case involving a potential international relocation

64

CC-55, [79]; similarly CC-27, [A31]; CC-41, [5]; CC-94, [5]. 65

CC-62, [34]. 66

CC-69, [98]. 67

J. Herring and R. Taylor, ‘Relocation Relocation’ [2006] CFLQ 517. 68

YC v United Kingdom (Application No 4547/10) [2012] 2 FLR 332 (ECtHR). 69

Re B (Care Order: Proportionality: Criterion for Review) [2013] UKSC 33, [2013] 2 FLR 1075, [45]. 70

Re Y (Removal from Jurisdiction: Failure to Consider Family Segmentation) [2014] EWCA Civ 1287,

[2014] 3 FCR 483.

the potential for an interference with Article 8 rights means that a clear and separate

proportionality evaluation may be required. Quite what this involves is not spelt out by Ryder

LJ, but experience in the public law field suggests that a careful assessment involving the

parallel analysis of all realistic options can be done in a way which involves the

proportionality assessment and overall balancing exercise required under Article 8.71

Factors coming from Payne

The applicant’s proposals

Both judgments in Payne ask about the applicant’s proposals. Judges usually scrutinised the

substance of applicants’ relocation proposals in great detail. For example, in many cases

where relocation was refused, judges made critical remarks about the details of the plans.

While occasionally these criticisms were substantive,72

a more common criticism was that the

plans were ‘vague’,73

‘ill thought out and ... over-ambitious’,74

or ‘not well evidenced’.75

While, of course, these remarks may have been entirely justified, a plan that one person sees

as ‘over-ambitious’ could be considered exciting and potentially highly rewarding by another.

For example, in another case the judge said of the mother:

I accept that she may not have dotted every ‘i’ and crossed very ‘t’, but frankly ... the

mother seems to me to be so capable that any unexpected obstacle could – and would – be

easily overcome by her.76

In the end, while judges spoke at greater length about practicalities, their views about the

each parent’s truthfulness and reliability often seemed more influential.

The broad pattern was that unsuccessful parents were the subject of criticism from judges

about what they had claimed or the way in which they had claimed it. For example, in about

half the cases where relocation was refused, judges were concerned about the applicant

parent’s attitude to contact and her trustworthiness in relation to the child’s on-going

relationship with the father. One or two cases appeared to turn entirely on these findings,

such as CC-55, an application to move with children aged 8 and 7 to the United States. The

judge found that the ‘cunning and manipulative’ mother ‘seeks to undermine contact’; her

‘very abusive’ comments about the father were ‘indicative of the depth of hatred’ she felt.77

71

The European Court of Human Rights sometimes takes the same view: see, eg, YC v United Kingdom

(Application No 4547/10) [2012] 2 FLR 332 (ECtHR). 72

For example, CC-13 involved the mother proposing to leave her job in London and go to Canada, where

she had no other connections and where she would commute for nearly 3 hours each day for 4 hours of

low-paid work which would barely cover the costs of childcare. 73

CC-90, judge’s cover note. 74

CC-55, [87]. 75

CC-39, [5]. 76

CC-67, [20]. 77

CC-55, [77], [51], and [70] respectively.

On the other hand, cases where relocation was allowed were characterised by adverse

remarks about the respondent parent’s attitude to the child or to the applicant parent,78

often

coupled with negative comments about his evidence at the hearing. The latter of these ranged

from mild remarks about the father ‘not being completely frank’,79

through more significant

findings that the father was ‘a dishonest man’,80

and giving evidence which was ‘in parts

excessive and in parts false’.81

The effect of relocation on the child’s relationship with the non-moving parent

Payne next asks about the effect of a relocation on the non-moving parent, and about the

steps that can be taken to offset the loss of relationship with the child, about which judges

expressed a variety of views. With the exception of those handful of cases where the fathers

represented a safety risk to the children, judgments almost invariably spoke of there being ‘a

very loving and close relationship’,82

where ‘it goes without saying that this child has a strong

bond with her father’.83

Indeed, so similar were the comments across the body of cases that,

looking just at these sections of the judgments, it would be difficult to differentiate between

those applications that were granted and those that were refused. A similar convergence of

language is seen in those cases where judges spoke about the loss of the ‘everyday parenting

experience’ – what Eleanor King J once described as ‘lying around on a sofa on a Saturday

evening, eating pizza and watching a DVD with your dad, or being taken to school by him

every other Monday’.84

By contrast, a small number of cases proceeded on a basis which seemed to lack much

appreciation of the importance of these ordinary everyday interactions. Case CC-44 was a

particularly egregious example, where the child had been mainly looked after by the father on

a day-to-day basis for a year. Allowing the mother to remove the child from the father’s care

to join her abroad, the judge said this:

Furthermore, and I know that the father will find it difficult to accept this, it seems to me

that [the child] will have a better relationship with his father in many ways if regular

contact takes place. I did get the impression ... that an awful lot of time that [the child]

spent with his father was either at school or being fed or doing homework and not much

of it was time for [the child] and his father to have fun together ... in what one might call

78

For example, a drug-dealer father who had been in prison for most of the child’s life claimed ‘that he had

been a good role model to [his son, which] demonstrated ... a lack of understanding of the impact of his

lifestyle on his then-family’: CC-62, [10]. 79

CC-81, [31]. 80

CC-5, [28]. 81

CC-12, [29]. 82

CC-72, [10]. 83

CC-7, [82]. 84

J v S (Leave to Remove) [2010] EWHC 2098 (Fam), [2011] 1 FLR 1694, [99].

down time. This will be available to his father on contact.85

Many would find this a truly extraordinary statement. Of course a large part of the week for a

school-aged child involves being at school, having dinner and doing homework. The father

was continuing to work full-time in his job, since that was the family’s only financial support,

and weekends were often being spent having contact with the mother. To tell the father that

losing the day-to-day care of his son would lead to ‘a better relationship’ was, at best,

unfortunate language.86

Here we see direct engagement by the judge with the father, but the

sexist undertone and detachment from the father’s own experiences as claimed in evidence

are likely to have undermined the intended positive effect.

Much the same sentiment can be found in case CC-56, though the case did not involve the

same direct engagement between judge and parent. Here, the judge accepted that the father

had been providing 65% of the children’s care for the three years since the parents’

separation. He had given up full-time work to do so, while the mother continued in her

demanding job in London and started a relationship with a man (to whom she was now

married) who lived in the USA. As the judge said, ‘[m]ainly her career but also her current

marital choice have impacted upon her ability physically to spend time with [the children]

though she has done her utmost’.87

Somehow, the fact that ‘a very high proportion of [the

mother’s] contacts have been at extended weekends and thus there was proportionately more

time available with the children’88

counted for more than the father’s daily commitment,

because the mother was said to be providing a ‘pivotal emotional relationship’ such that ‘the

welfare of the children is threatened by [her] inability to cope and thereafter to provide

emotionally’.89

Despite the judge’s recognition that ‘[the father] undoubtedly loves his

children and has been a good father to them, shouldering as he has the greater bulk of the

parenting duties’,90

still the inherent qualities of the mother – who might alternatively have

been thought to have behaved selfishly and without thought for the children – made her more

suitable to care for them and justified a removal from the father’s care and from the

jurisdiction.

Where the judge allowed relocation, there was usually a statement to the effect that this

father-child bond was strong enough to survive the resulting geographic separation. As one

judge put it, ‘the fundamental strength of the relationship ... would not be impaired’.91

In

another case, the judge thought that ‘[t]he quantity of contact will be reduced but the quality

85

CC-40, [26]. 86

See also CC-48, [111]: ‘I am sure [the children] will miss [the father] and he will miss them, but this will

make their regular reunions all the more satisfying for them.’ 87

CC-56, [24(ii)(b)]. 88

Ibid, [23]. 89

Ibid, [47]. 90

Ibid, [53]. 91

CC-48, [111].

will not in my judgment be affected’.92

Judges had ‘no doubt that the father will be

devastated if permission to relocate is given’93

and endeavoured to make orders for on-going

contact that reflected the fact that ‘[the] father remains hugely important ... and he will see

[the children] just as often and for as much time as is possible’.94

By contrast, a number of judgments in which relocation was refused expressed concern that

the relocation might risk a loss of the meaningful relationship between the child and the

father. In some cases, these concerns were based on a lack of confidence in the mother’s

commitment to contact, but more often the perceived risk was based on a view that the

distance involved and the consequent change in the nature of the relationship would be too

serious for a meaningful relationship to survive.

A majority of cases where this loss or damage was seen to be a ‘decisive point’ in refusing

relocation involved an existing strong relationship between child and father,95

but in others it

was thought that the child’s relationship with the father needed strengthening, and that

relocation would prevent this from taking place; here, the relocation was prevented largely in

order to improve a currently poor relationship.96

These two classes of comment might suggest certain judicial (or legal) assumptions about the

benefits of a strong father-child relationship.97

Relocation might be refused because of an

existing strong relationship which was already important to the child, or might be refused

because there was no such relationship and it needed to be built and strengthened.

Conversely, relocation might be allowed because the parent-child bond was strong and would

allow the relationship to survive the distance, or it might be allowed because the bond was

not that strong and so the risk of loss was less important because the relationship was not as

central to the child’s welfare as other factors.

These judicial choices explicitly relate to the law’s present uncertainty about parenting norms

and the importance of fathers in children’s upbringing. In 2014, after the cases in this study

were decided, the Children Act 1989 was amended to include a so-called ‘parental

involvement presumption’.98

This provision states, in short, that a parent should be involved

in his or her child’s life unless it is not safe for that to happen, but that ‘involvement’ does not

mean any particular division of time. While this new subsection has not yet been the subject

of significant judicial consideration, either in general or in the relocation context specifically,

Ryder LJ has said that in an international relocation case s 1(2A) is likely to ‘heighten the

92

CC-12, [35]. 93

CC-96, [14]. 94

CC-61, [39]. 95

For example, CC-30 and CC-84. 96

CC-36 and CC-55. 97

See, eg, R. Bailey-Harris, J. Barron and J. Pearce, ‘From Utility to Rights? The Presumption of Contact in

Practice’ [1999] IJLPF 111. 98

Children Act 1989, s 1(2A), inserted by the Children and Families Act 2014.

court’s scrutiny of the arrangements that are proposed by each parent’.99

Effect of refusal on the applicant

Set against this, judges considered the likely impact that a refusal of the application might

have, where again judges seemed to be able to construct identical arguments leading to

opposite conclusions. The issue of the effect of refusal is, perhaps, the area which most tests

trial judges in the relocation context. In the wake of several Court of Appeal judgments in the

early 2000s,100

a narrative was established that judges were required to find that there would

be a substantial impact of refusal and that this factor would carry substantial weight in the

overall analysis.101

While recent decisions have stressed that this is not the case,102

in an area

like relocation, where most judges are inexperienced and may see a case only once every two

or three years, it takes time for these changes to filter down. Consequently, judges may still

see it as ‘safer’ to allocate considerable weight to the impact of refusal, while a finding that

there would be little or no impact – or, even more so, that there would be impact but that it is

outweighed in the overall evaluation – remains ‘braver’.103

It is therefore noteworthy to find many judges pointing out – in line with the more recent

authority104

– that the effect on the parent of refusing relocation was only one of many factors

to be assessed based on the evidence in the case:

I give no primacy to the mother’s application and do not adopt any approach which

creates some presumption in her favour by attaching weight to the emotional and

psychological impact upon the mother, or any consequential impact upon the child. These

factors do, nevertheless, have to be considered as part of the overall balance.105

In assessing a possible shift away from the one-time focus on the effect of refusal of leave,

two trends can be noted. One, in the cases where relocation was refused, involves judges

willing to find either that the impact of refusal would not be as bad as the mother thought, or

that she was able to deal with that effect well enough.

[A]lthough the mother would be very disappointed ... she would cope and ... she would

not allow her disappointment to taint the loving relationship that either parent has with

99

Re F (International Relocation Cases), [35]. 100

In particular, Re B (Removal from Jurisdiction); Re S (Removal from Jurisdiction) [2003] EWCA Civ

1149, [2003] 2 FLR 1043 and Re B (Leave to Remove: Impact of Refusal) [2004] EWCA Civ 956, [2005]

2 FLR 239. 101

M. Hayes, ‘Relocation Cases: Is the Court of Appeal Applying the Correct Principles?’ [2006] CFLQ

351, 366. 102

Re W (Relocation: Removal Outside Jurisdiction) [2011] EWCA Civ 345, [2011] 2 FLR 409; K v K. 103

Relocation Disputes, p 79. 104

At the time, the clearest expositions were J v S (Leave to Remove) [2010] EWHC 2098 (Fam), [2011] 1

FLR 1694 and K v K. The point is reiterated by the general requirement of a holistic, non-presumptive

approach from Re F (International Relocation Cases) [2015] EWCA Civ 882. 105

CC-21, [29].

the child. She would move on. Indeed, she said so herself in oral evidence.106

[T]he mother will undoubtedly be extremely upset... [T]he Cafcass officer was of the

clear view that the mother would cope ... [and] I formed the same view. ... I feel sure that

the mother will cope in this country, as she has done for many years.107

These cases – including some which seemed to involve judges instructing mothers that they

were required to cope108

– illustrate the willingness of judges to find that the overall balance

of the welfare assessment favours refusal of relocation, despite the effect on the applicant.

The competing trend, where relocation was allowed, shows that a successful application can

be brought without any reliance on the effect of refusal. In a small but substantial minority of

cases, ‘the mother’s case is not put on the basis that she will suffer severely if not permitted

to travel’.109

As another judge explained:

This is not the kind of relocation case in which a mother states that she cannot continue

with life in England... When asked how she would feel about the refusal of her

application, the mother in effect said that she would just have to cope with it. ... In fact,

unusually for an applicant, her situation in England is, I believe, considerably tougher

than she admits.110

Set against these cases were a sizeable number of applications which were allowed where

judges clearly considered that the impact of refusal was a major consideration. In many of

these, the judges linked their conclusion to specific parts of the evidence:

If she was not allowed to go the nightmare ... that her life ha[s] become would continue,

and I do not think that is an unfair way of describing it.111

Mother was trying to hide how distressed she feels at her situation here... Her desire to go

home and her devastation if she has to stay here were plainly genuine. ... I do think she

will be devastated by a refusal. She was trying to be brave ... [but] she is trapped and

worn down by life here.112

If refused permission she would be very distressed and frustrated and would entirely

blame the father. ... She would be badly affected if refused permission, with an indirect

effect on [the child].113

106

CC-67, [21]. 107

CC-85, [87(e)]. 108

For example, CC-30, [81]. 109

CC-14, [41(5)]. 110

CC-27, [A51]. 111

CC-5, [40]. 112

CC-49, [40] and [56] respectively. 113

CC-62, [25] and [30].

However, in other cases, the conclusion that the mother would be ‘devastated’ seemed less

compelling. In one case, the parents were operating a shared care arrangement in central

London and the mother’s job at a large multinational company took her to Brussels on a

regular basis; she proposed to move there full time, though her employer was content to have

her based either in Brussels or in London. The judge’s view in this case that the mother

‘would be extremely upset, indeed devastated, if she could not move to Brussels’114

seems,

with respect, to overstate the case. The mother, irritated no doubt, would in all likelihood

have carried on her life much as before, and perhaps in time sought a posting that required

less travel (surely a possibility, given that she worked for a major international company with

its global headquarters in London).

This point was well made by another judge, who observed that ‘hyperbole has become

jargon’115

at the start of eight full paragraphs dealing with the ‘devastation’ issue:

Both counsel, and I am bound to say I, sighed when we heard the word ‘devastated’ yet

again. The problem is that there is a limited vocabulary of words available to describe the

extent of a person’s feelings of despair, loss, grief, fear, exasperation, frustration and

disappointment, but in so saying I illustrate that there are other words than ‘devastated’.

In the mother’s case, she told the father that the refusal of her wish to relocate would be a

disaster. She would lose her identify, autonomy and independence; and, inevitably, she

would be devastated. I found that to be a considerable overstatement of what is likely to

be the truth.116

That is not to say that the judge discounted the effect on her – it was considered and weighed,

and the application was allowed. This judge was not alone, however, in expressing concern at

the dominance of claims of ‘devastation’. Another judge, who also allowed the relocation

application, found expressly that ‘M[other] will feel distress – but not devastation, as she said

in evidence – if she does not [relocate]’.117

A number of cases where relocation was

ultimately refused also commented on this point, with the word ‘disappointed’ featuring often

as a likely impact; in one particularly telling case the maternal grandmother, giving evidence,

‘thought her daughter would be devastated if leave were refused, but conceded that the word

“devastated” was suggested to her by the mother’s lawyer’.118

Children’s wishes and feelings

An issue not mentioned expressly in the Payne guidance, though imported into it by its

114

CC-81, [79]. 115

CC-48, [112]. 116

Ibid. 117

CC-72, [17]. See also CC-67, [21]: ‘the mother would be disappointed with refusal, but I hesitate to use

the word “devastated” which was employed during the hearing as this seems to me to be putting it far too

high’; CC-73, [65]: ‘I accept that the mother would be upset and distressed ... but ... I do not accept that

she would be devastated’. 118

CC-36, [28].

reference to the welfare checklist, is the views, wishes and feelings of the children involved.

Such wishes and feelings are principally brought into the family justice system through an

officer from the Child and Family Court Advisory and Support Service (Cafcass) or an

Independent Social Worker’s report. Such reports can be limited to the issue of the child’s

wishes and feelings, though they tend to be much more wide-ranging and encompass all

aspects of the case. A full welfare report typically carries great weight in parenting disputes;

if the report includes a recommendation, judges are required to give specific reasons for

rejecting that view if they reach a different conclusion.119

However, it has long been thought

that Cafcass reports are less influential in relocation cases than elsewhere in child law.120

Indeed, in the cases in this study, reports were not invariably ordered, often because the

parties and the judge simply felt they did not need one.

Where Cafcass reports were made, judicial reactions were mixed. One area in which judges

were generally united, though, was that the report was important in relaying the wishes and

feelings of the children involved. However, in practical terms, children’s views had quite

limited impact in the judges’ reasoning in most cases. Since the mean age of children in

relocation cases is just under 7 years, it is perhaps unsurprising that judges often simply noted

that the children were ‘too young for [their] wishes to assist me’,121

or even that they were

‘too young to express [their] wishes and feelings on the issues before the court’.122

For children in middle childhood (roughly ages 5 to 10), the most common approach was to

say that the issues were too complex for their wishes and feelings to carry much weight.

While one may have some sympathy with this view, it should be remembered that the

children were not being asked to decide the case, but rather to express their wishes and

feelings about it. While these children may not ‘realise fully what would be involved’ in

relocation,123

some showed real insight into what was being decided. One 7 year old who had

been living in England for about three years rather wisely said to the Cafcass officer that ‘he

missed his life in [the other country] but that if he was [there] he might just as well have been

saying that he missed his life in England’.124

The judge thought that these expressed views

‘don’t help me very much’ in making the decision, but one might have inferred that the child

was aware of the benefits of his present life and was not expressing any desire to change.125

Indeed, in other cases, children of this age did express views which influenced the decision-

making. Usually these were cases where the children’s views accorded with the outcome that

119

See, for example, Re V (Residence: Review) [1995] 2 FLR 1010 (CA). 120

Relocation Disputes, p. 103. 121

CC-41, [41]. 122

CC-74, [69]; also CC-53, [47]. 123

CC-39, [14]. 124

CC-40, [10]; also CC-72, [25], where another 7 year old had written two letters to the judge which

revealed ‘that [the child] is very torn indeed’. 125

CC-40, [25]; see also CC-68, where the 11 year old child made the seemingly important point that she

had never even visited the proposed destination country and ‘thought she should go on holiday there first

to see if she liked it’ (para [4]).

the judge had reached, such as Case CC-69, where the parents were in a high-conflict shared

care arrangement which the judge thought was not working well. The judge reported this

exchange when the 8 year old met the judge in person:

I asked him ‘is there anything you want to say?’ And he said ‘I want to live with my

mum’ and because he said that I felt I had to ask him ‘why?’ He said ‘when I see my

mum my tummy doesn’t ache any more, and when I see dad my tummy aches’.126

Unsurprisingly, in the handful of cases involving older children judges often said that their

views ‘are something I have to give a lot of weight, albeit they are not determinative’.127

One

13 year old spoke for many children when reporting that ‘she would, herself, [have] preferred

not to have been asked’.128

Despite that, the judge found significant value in what the child

said:

she told [the Independent Social Worker] that she would be fine either way, [and]

expanded that what she really wanted for herself (and I underline this for emphasis) was

to move to the United States and go to school there, but she would miss her dad and

would prefer it if dad moved there too and lived nearby.129

Thus case outcomes generally accorded with older children’s wishes, but there were

exceptions. In one case, relocation was granted in relation to a 13 year old boy who was

described as ‘very apprehensive’ about the proposed relocation and who was ‘close to tears’

when discussing it with the Cafcass officer.130

Two cases involved the judges specifically

discounting the views of older children, one because ‘she tells people what they want to hear

– she is a people-pleaser’,131

and the other because the judge concluded that the children’s

views had been unduly influenced by their mother and were therefore simply expressing her

views rather than their own.132

Overall, it is clear that in only a small minority of cases did the children’s views have any

great influence on the decision. In a high proportion of cases their views were either not

sought or were entirely dismissed – the child was ‘too young’ or their views were discounted

for some other reason – and even in those cases where weight was being given to the child’s

views, the impression from the judgments was that the decision would have been the same

regardless of the child’s view. In other words, judges said that they were giving weight to the

child’s views when those views aligned with their own conclusions about the outcome of the

126

CC-69, [76]. Note that judges are strongly discouraged from obtaining evidence directly from children:

Family Justice Council Guidelines for Judges Meeting Children subject to Family Proceedings 2010; Re

KP (Abduction: Judge Meeting Child: Conduct of Interview) [2014] EWCA Civ 554. 127

CC-64, [17]. 128

CC-48, [125]. 129

Ibid, [124]. 130

CC-33, [6] and [26] respectively. 131

CC-43, [13]. 132

CC-85, [75].

case (a reinforcement approach), and found reasons to dismiss the child’s views when the

outcome was not going to fit with them.

CONCLUSIONS

This is the first study in England and Wales that gives any insight into the outcomes of

international relocation cases in the family court. While the two parts of the study show

slightly varying results in places, the overall picture is clear and informative. We see, most

crucially, that it cannot be said that relocation is automatically allowed – in this sample at

least, something like 30 to 40% of applications were refused. We can also identify some

variables which, overall, may be related to those case outcomes, such as the distance of the

proposed move and the applicant’s current relationship status.

Moreover, we have been able to look in detail at the reasoning that goes into making a

determination in a relocation case. Presented with enormous amounts of evidence on wide-

ranging issues, trial judges have many choices to make. They have to decide what is relevant

at all and in what ways it is relevant; they have to choose between competing (often directly

contradictory) versions of the same events; and they have to make predictions about various

possible futures for a child, with effects potentially lasting a lifetime. They then have to

construct a narrative account of their choices which weaves together their findings and

conclusions within the legal framework so as to allow the decision to be understandable,

coherent and (ideally) with minimal scope for appeal.

The research reported in this article tells us a great deal about the decision-making processes

of judges in everyday relocation cases. We can see, in particular, that judges give very serious

consideration to two main competing factors – the effect on the applicant of refusing her

application, and the effect on the child’s relationship with the respondent if relocation goes

ahead. These are assessed against the background considerations of the practicalities of the

proposal (including considerable focus on the plans for on-going contact), and the

motivations of both parents.

In most cases, judgments were constructed such that all these aspects lined up to point

towards the same conclusion. Sometimes that was, no doubt, an accurate reflection of the

way in which the case had developed, but other cases raised some doubts, given that the same

issue could be constructed to lead to exactly opposite conclusions: for example, a strong

existing relationship between the child and the respondent could be seen as a reason to refuse

a relocation (more to lose) or to allow one (relationship strong enough to survive the move).

By contrast, a handful of judgments expressed genuine anxiety and uncertainty, which may

more accurately reflect the reality of many relocation cases.

More generally, the big question about relocation when these cases were decided was

whether the Court of Appeal’s reinterpretation of Payne in K v K went far enough,133

or

133

This is Mostyn J’s current view, according to NV v OV [2014] EWHC 4130 (Fam), [7].

whether the guidance itself needed to be rewritten. This research suggests a rather mixed

answer to that question.

A number of concerns were expressed about the ‘Payne in the light of K v K’ approach.

Perhaps the main concern was that Payne was premised on there being a primary carer

applicant, when many cases no longer fit that pattern. K v K (as understood after Re F134

)

demands that judges deal with this issue by weighing the Payne considerations differently,

but there was a question as to whether Payne lends itself to such flexible interpretation.135

Here, the findings of this research are somewhat equivocal. Some judges clearly understood

that they were permitted to take this approach and did so with little apparent difficulty. In

other cases, though, it appeared that Payne was applied fairly uncritically, despite being

inappropriate to the facts of the case.136

Whether because of greater seniority, greater

experience, less risk-aversion,137

or some other consideration, some judges readily modified

the Payne guidance while others followed it closely as if it were a manual. This, in short, was

the difficulty: there was a concern that K v K risked an inconsistency of approach and

therefore of like cases not being treated alike. Moreover, as relocation cases are increasingly

allocated to the District Bench, the Court of Appeal’s ability to have an overview of the

general approach diminishes, since most applications for permission to appeal will now be

heard by one of dozens of Circuit Judges rather than the handful of specialists in the Court of

Appeal.

It may be that the Court of Appeal’s decision in Re F (International Relocation Cases) in

August 2015 has changed the picture somewhat. Here, Ryder LJ suggested that undue

reliance on Payne’s guidance could amount to an appealable error of law, and that the

‘required reading’ for a judge in an international relocation case no longer included Payne,

but instead came principally from K v K and Re F (Relocation). Even in the lifetime of the

cases in this study, Re F (Relocation) appeared to be helping to standardise understandings of

the approach after K v K, and it may be that Re F (International Relocation Cases) goes even

further in standardising a holistic, all-factor approach.

Nonetheless, it may be wondered whether Payne will continue to exert influence, since it has

still not been over-ruled. K v K, which Ryder LJ posits as the central case, is itself focused on

the continued relevance of Payne, and thus working back through the authorities one still

ends up with the guidance from that case. The concern, looking at the reasoning of judges in

this study, is not that the issues raised by Payne are inappropriate in many cases; rather, it is

the manner in which those questions are asked and the assumptions which underpin them

(particularly in relation to the child having a single primary carer who is inevitably the

134

Re F (Relocation) [2012] EWCA Civ 1364, [2013] 1 FLR 645. 135

R. George, ‘International Relocation, Care Arrangements and Case Taxonomy’ [2012] Family Law 1478;

E. Devereux and R. George, ‘When Will the Supreme Court Put Us Out of Our Payne?’ [2014] Family

Law 1586. 136

CC-40 and CC-56 were the most blatant examples. 137

On judges’ concerns about appeals, see Relocation Disputes, p 79

mother) which seem inappropriate, and asking judges to determine the extent to which that

guidance is relevant on a case-by-case basis may be too much of a stretch.

Indeed, it is difficult to see why the senior judiciary remains so attached to Payne. It cannot

really be about precedent, because Payne is only guidance and the Court of Appeal has

regularly abandoned one set of guidance in favour of another – that, indeed, is how Payne

became the leading case in the first place. Despite the clarification from Re F (International

Relocation Cases), it is time for the Court of Appeal or the Supreme Court to look at this

issue again from scratch, ideally by taking a number of conjoined appeals, with the aim of

issuing new guidance that applies to all cases without relying on individual judges to make ad

hoc modifications.138

138

My proposed guidelines are set out in Relocation Disputes, pp 157-9.

METHODOLOGICAL APPENDIX

This appendix describes the way in which the dataset was generated, and the resulting

limitations of the data in terms of their representativeness or otherwise of relocation cases as

a whole which reduce the extent to which they may be relied upon to make inferential

conclusions about this class of case generally based on statistical significance of any of the

findings.

The Court Case Sample

The CC sample of cases was decided in the calendar year 2012 and gathered directly from

trial judges and practitioners.139

Judges were contacted directly by email and letter. Initial

contacts were made by Thorpe LJ, with follow-up communication through the Family

Division Liaison Judges and the Designated Family Judges. Individual judges were also

contacted personally where I received a ‘tip off’ that they had heard a relocation case

recently. Individual judges were very helpful in sending materials when asked, but of course

it is not possible to know how many decided a case but did not remember to send it in, or

chose not to do so. Following a low response rate in the early part of the study, lawyers were

then contacted in various ways to ask for their assistance, once it was clear that the

President’s authorisation for the project allowed anyone who had a judgment or order to

submit it to the project.

The CC study was an attempt to gather the entire population of relocation cases heard in

England and Wales that year. It is not suggested that this attempt was successful;140

96

eligible cases were collected,141

but it is difficult to assess what proportion of the total

population of relocation cases this is. Official records give some insight into how many

international relocation cases are heard by the courts each year. Ministry of Justice statistics

show that, in 2012, 384 children were the subject of orders allowing them to be removed

from the UK following private law proceedings.142

However, caution may be needed in

approaching this number. First, it is not clear how the MoJ data are collected, and so the

likely accuracy of the number is difficult to assess. Second, it relates to the number of

children, not the number of cases. There were 1.5 children per case in the CC sample which,

if typical, would suggest something like 260 cases in total. Third, the MoJ data do not appear

to differentiate between permanent and temporary leave to remove a child from the

jurisdiction. Nonetheless, this number gives some insight into the size of the CC sample

139

Sir Nicholas Wall, President of the Family Division when this project was running, authorised access to

these cases. 140

Because both judges and lawyers were being asked for cases, a good many of the 96 were collected more

than once. Many judges also reported, directly or through Thorpe LJ’s office, that they had not heard any

relocation cases in 2012. 141

A number of other cases with international elements were sent in, but discarded as not being relocation

cases within the definition used in this study. 142

See http://www.justice.gov.uk/downloads/statistics/courts-and-sentencing/csq-q4-2012/csq-q4-2012-

family.csv, column DR.

relative to the total number of cases. Assuming the figure of 384 children to be correct, and

assuming 1.5 children per case on average, and assuming that holiday cases might amount to

between 0 and 20% of those cases (though the number is unknown), the 96 cases in the CC

sample would represent between 36 and 46% of cases. It is not possible to know whether the

cases are representative or not, though having both lawyers and judges sending cases goes

some way towards mitigating the risk of intentional bias on the part of those submitting

material, since the incentives for the two professional groups to do so may differ. While a

handful of judges and lawyers sent more than one case, the vast majority of individuals

sending material sent just one case, and some cases were received more than once from

different sources.

The Research Questionnaire Sample

In the final three months of 2012, a research questionnaire was distributed to family law

practitioners throughout England and Wales. The questionnaire was intended to provide an

alternative source of data about cases decided in 2012, to triangulate findings against the CC

sample and check for possible biases in the case sample submitted to the study. It was also

useful for bolstering the overall number of cases being considered, though both direct

collection of cases for the CC sample and the questionnaire study for the RQ sample faced

significant difficulties in obtaining responses from busy professionals.

Most questions asked lawyers to provide detailed information about their most recent

relocation case, designed to elicit information comparable with the data collected from the

court cases themselves. Practitioners were asked to complete a questionnaire about their most

recent relocation case, international or domestic, if they had done one in the previous 3 years

(and also to send nil returns), and the questions then differentiated between type of case, year

of decision, and mode of resolution (court adjudication or settlement). Some aspects of the

RQ data suggest that some respondents may have reported their most recent ‘interesting’ case

(however they saw that), since – for example – the sample appears in places to be skewed

towards High Court decisions.

The questionnaire was prepared in hard copy and online. 1,000 hard copies were distributed

at practitioner conferences and by direct mail to chambers and family law firms around

England and Wales, selected to get a range or large and small firms / chambers in all areas of

the country. The online version was also advertised in articles and news items posted in

practitioner journals and websites, and through the practitioner organisations Resolution, the

FLBA, and the ALC. 187 completed responses were received, including 8 nil returns. While

this number suggests a low response rate, it should be borne in mind that relocation cases are

not everyday work for family lawyers, and a great many practitioners will not have done one

in the last three years, even if they do private law children work. Of the 187 total responses,

52 related to judicially determined international cases heard in 2012, and it is these 52 cases

which are used in this paper, being directly comparable with the 96 CC cases.

Data Analysis

This paper draws on a mixture of quantitative and qualitative methods. For the purpose of the

quantitative analysis, each case was entered into a database containing 84 fields, though no

single case had data relevant to every field. These fields included basic information such as

the number and ages of the children, the proposed destination country and the case outcome,

but also more detailed information about process (such as location of court, type of judge,

whether each party had a lawyer), the family (such nationality of parents, presence of wider

maternal and/or paternal family in this country and in the proposed destination), and the

outcome (such as the extent of contact ordered). Many fields simply did not apply to many

cases (such as whether there were child protection concerns). But particularly where only an

order (i.e. no judgment) was available,143

there were cases where more common fields could

not be completed (such as the parents’ nationalities). Consequently, when it comes to analysis

of the data, it is not always possible to consider all cases in relation to particular variables.

The quantitative data were then analysed by Dr Ornella Cominetti, a statistician then at the

Oxford University Department of Statistics. She used three primary techniques to assess the

data: Pearson’s chi-square test, logistical regression models, and information gain score (IGS)

analysis. While the first two are well known, IGS is a less common technique. IGS evaluates

the worth of each variable by measuring the information gain of a variable to the class

(where, in this research, the outcome variable is whether the relocation is allowed or not).144

The qualitative analysis related to the 53 cases from the CC sample in which court judgments

were available. The judgments were analysed using a broadly grounded methods theory, but

one informed by the fact that many of the key issues being sought were known in advance

based on the research questions – since the questions were partly about the use that is made

of the Payne guidance, issues identified within that guidance formed part of the coding

framework from the start.

143

Some questionnaires also had questions left blank, or respondents had written ‘don’t know’ next to a

question. 144

See M. Hall and G. Holmes, ‘Benchmarking Attribute Selection Techniques for Discrete Class Data

Mining’ (2003) 15 IEEE Transactions on Knowledge and Data Engineering 1437, 1447.